Brazilian Election Faces AI Misinformation Crisis

It is common for AI videos containing promises candidates never made to spread as truth, leading residents to believe them and experience confusion in their political judgment.

In the Brazilian presidential campaign, AI has infiltrated election efforts, with hundreds of videos creatively produced and competitively distributed, mainly by the right-wing.

For example, AI-generated videos linking President Lula to Nicolás Maduro or depicting candidate Jair Bolsonaro as a savior have been circulated in attempts to manipulate voters' perceptions.

Propaganda featuring a fictional elderly character like Dona Maria has emerged to attack voters, and some popular AI videos have garnered over 9 million views, demonstrating their powerful reach.

Creators argue that using AI for satirical purposes is legal, but in reality, despite legal requirements for candidate consent and disclosure, compliance is extremely difficult.

While technology to detect deepfakes by analyzing facial movements and image noise is advancing, there are limitations to completely preventing actual election interference.

Experts analyze that over 50% of AI videos are for promoting specific candidates, 25% are for satire, and less than 25% are for slandering opponents. They warn that customized deepfakes targeting 40-45% of swing voters could change election outcomes.